1345
Nu Hosp. 2015;31(3):1345-1351
ISSN 0212-1611 • CODEN NUHOEQ
S.V.R. 318
O iginal / Valo ación nu icional
Can we imp o e he bi h weigh p edic ion? he e ec o no mal BMI
using a mul i a ia e model
R. Vila-Candel1, J. M. Ma in-Mo eno2, S. Alama 3, F. J. So iano-Vidal4, F. G. Na anjo de la Pue a5 and
M. Mu illo6
1PhD, Midwi e y, Hospi al Uni e si a io de la Ribe a, Spain. Di ec o o Depa men o Nu sing, Uni e sidad Ca ólica de
Valencia. 2MD, PhD, D PH is a Full P o esso a he Depa men o P e en i e Medicine and Public Heal h, Uni e si a de
Valencia, Spain. Di ec o , P og amme Managemen Wo ld Heal h O ganiza ion-EUROPE. 3PhD, Senio Nu sing Lec u e ,
Uni e sidad Ca ólica de Valencia, Spain. 4Midwi e y, Hospi al LLuis Alcanyis, Spain. Senio Nu sing Lec u e , Uni e sidad
Ca ólica de Valencia. 5MD, PhD, Di ec o o Depa men o Obs e ic and Gynaecology, Hospi al Uni e si a io de la Ribe a,
Spain. Associa e P o esso a Uni e sidad Ca ólica de Valencia. 6Nu sing, MSc, Senio Nu sing Lec u e , Uni e sidad Ca ólica
de Valencia, Spain.
Abs ac
Objec i e: The cons uc ion o a p edic i e model ha
imp o es he es ima ion o he e al weigh (EFW).
S udy Design: a compa a i e, desc ip i e s udy.
One hund ed o y p egnan women we e ec ui ed a
wo-s age sample in heal h depa men in Spain. They
we e classi ied in ou g oups depending on he p e-ges-
a ional BMI. Fe al weigh a e m was es ima ed by ul-
asound a 33-35 weeks (EFW40w) by one gynecologis .
A eg ession model was c ea ed wi h he a iables ha
eac ed o he newbo n´s weigh , symphysis- undal hei-
gh (SFH), EFW40w, ges a ional age (GA), e i in le el
and ciga e es smoked.
Resul s: A mul i a ia e model was c ea ed o he NW
g oup o es ima e he e al weigh (EFWme), esul ing
in R2=0.727 (p<0.001). The di e ences o he a e ages
ob ained be ween EFW40w and EFWme, wi h he new-
bo n´s weigh we e signi ican (p<0.001). EFWme unde-
es ima es bi h weigh by 0.07 g (mean e o 0.53%),
and EFW40w o e es ima es i by 300.89 g (mean e o
10.12%). In o de o e alua e he p edic i e model and
e i y he p edic ions we used he Bland-Al man analy-
sis. The a e age e o in es ima ing he bi h weigh wi h
EFWme was 1.94% unde es ima ing he esul , whe eas
he ul asound e o o e es ima ed he esul 10.93%.
Conclusion: The mul i a ia e model c ea ed o he
NW g oup imp o es he accu acy o he ul asound.
(Nu Hosp. 2015;31:1345-1351)
DOI:10.3305/nh.2015.31.3.8150
Key wo ds: Bi h weigh . P egnancy. Ul asound. An h o-
pome y. Mul i a ia e analysis.
¿PODEMOS MEJORAR LA PREDICCIÓN
DEL PESO AL NACER? EL EFECTO DEL IMC
PREGESTACIONAL NORMAL USANDO UN
MODELO MULTIVARIANTE
Resumen
Objec i o: cons ui un modelo p edic i o que mejo e
la es imación del peso del ecién nacido (PFE).
Ma e ial y Mé odos: Es udio obse acional dónde 140
ges an es ue on es udiadas median e un mues eo bie-
ápico en un Depa amen o de Salud en España. Fue on
clasi icadas en cua o g upos dependiendo del IMC p e-
ges acional ma e no. El peso p oyec ado al nace ue es-
imado po la ecog a ía ealizada en e las 33-35 semanas
de ges ación (PP40s). Se cons uyó un modelo de eg e-
sión con las a iables que se eaccionaban con el peso al
nace , al u a u e ina (AU), PP40s, edad ges acional (EG),
ni el de e i ina y ciga illos consumidos.
Resul ados: Se cons uyó un modelo mul i a ian e
pa a el g upo No mo-peso pa a es ima el peso al nace
(PFm) ob eniendo una R2=0,727 (p<0,001). Las di e en-
cias de las medias ob enidas en e PP40s y PFm, con el
peso del ecién nacido, ue on signi ica i as (p<0,001). El
PFm in a alo a el peso al nace en 0,07 g (e o medio
0,53%) y la PP40s lo sob es ima en 300,89 g (e o medio
10,12%). Pa a e alua el modelo p edic i e y e i ica
las p edicciones ealizadas se u ilizó el análisis de Bland y
Al man. El e o medio de la es imación del peso al nace
median e modelo mul i a ian e ue de 1,94% in aes i-
mando el esul ado, en cambio el e o come ido po la
ecog a ía sob es ima el esul ado en 10,93%.
Conclusiones: El modelo mul i a ian e cons uido,
pa a el g upo NW, mejo a la p ecisión de la ecog a ía.
(Nu Hosp. 2015;31:1345-1351)
DOI:10.3305/nh.2015.31.3.8150
Palab as cla e: Peso al nace . Emba azo. Ecog a ía. An-
opome ía. Análisis mul i a ian e.
Co espondence: Ra ael Vila Candel
Hospi al Uni e si a io de la Ribe a.
C a Co be a, Km 1 Alzi a, 46600. Spain.
E-mail: [email p o ec ed]
Recibido: 30-IX-2014.
Acep ado: 3-XI-2014.
047_8150 ¿PODEMOS MEJORAR LA PREDICCIÓN DEL PESO AL NACER? EL EFECTO DEL IMC PREGESTACIONAL NORMAL USANDO UN MODELO MULTIVARIANTE.indd 1345 18/02/15 17:18
1346 Nu Hosp. 2015;31(3):1345-1351 R Vila-Candel e al.
In oduc ion
The analysis o bi h weigh mus be add essed om
a mul i ac o pe spec i e1. Un o una ely, bi h weigh is
unknown un il bi h akes place2 The use o ul asound
e al measu emen s ha e been ex ended and he measu-
emen s ha e been combined o es ima e e al weigh by
eg ession analysis o physically me hods3. Fe al weigh
es ima ion is inaccu a e, wi h poo sensi i i y o p edic-
ion a e m.4 I is al eady known ha he absolu e e o
a e age in p edic ing bi h weigh a ies om 6 o 12%
o he ac ual weigh . Se e al au ho s4-6 ha e shown ha
he le el o in a/in e obse e a iabili y in e al measu-
emen as well as he impac o e o s on g ow h assess-
men a e unaccep able. Di e en s udies7-9 ha e compa-
e, wi h disc epancies, he accu acy be ween clinical and
ul asound me hods in o de o es ima e e al weigh in
he hi d imes e . Bi h weigh depends on many ac-
o s i.e. ma e nal, gene ic and en i onmen al ones.10
This s udy aises he hypo hesis ha some ac o s
a e no dis ibu ed andomly, bu acco ding o a p o-
ile ha de e mines he weigh o a newbo n a bi h.
I could le us c ea e a be e p edic i e model o in-
an weigh a bi h, a he han he ac ual bi h weigh
es ima ion by hi d imes e ou ine sonog am, when
applied o women depending on hei p e-ges a ional
body mass index (BMI).
Ma e ial and me hods
Design
We pe o med an obse a ional and p ospec i e
s udy. Based on he WHO anges, p egnan women
we e alloca ed in ou di e en g oups depending on
hei p e-ges a ional BMI: unde weigh (UW <18.5
Kg/m2), no mal weigh (NW 18.5-24.9 Kg/m2),
o e weigh (OW 25.0-29.9 Kg/m2), obese (OB >30
Kg/m2). A sample o 159 p egnan women was collec-
ed om Feb ua y 2011 o Ma ch 2012.
A wo-s age sampling s udy was pe o med. In he
i s s age, wo su ge ies (Ca le and Benimodo) we e
chosen using a simple andom p obabili y sampling
om among all P ima y Ca e Cen es o La Ribe a
Heal h Depa men (Spain). In he second s age, p eg-
nan women we e selec ed using a p obabili y sam-
pling wi h andom s a and sys ema ic moni o ing
depending on he numbe o p egnancies pe yea ob-
ained in bo h o hem.
Inclusions c i e ia we e based on ma e nal age be-
ween 18 and 36 yea s, i s p ena al appoin men be-
ween 5 and 12 weeks o p egnancy and single- e us
p egnancy wi h no e al de o mi ies.
Exclusions c i e ia included e usal o pa icipa e in
he s udy, language ba ie , an ad e se obs e ic his o y
du ing p e ious p egnancies, medical condi ions ha
modi y e al g ow h, ma e nal in ec ion o any o he
ma e nal ch onic pa hology.
We es ima ed ha o a 95% Con idence In e al
(CI) and a 4% p ecision, we needed a minimum sam-
ple size o n=147.
E hical ques ions
This s udy was pe o med acco ding o he basic
p inciples o all medical esea ch se ou in he Decla-
a ion o Helsinki. The s udy was p e iously e alua ed
and app o ed by he Resea ch Commi ee o he Ribe-
a Uni e si y Hospi al.
Va iables
Six ca ego ies o a iables we e selec ed: an h opo-
me ic, demog aphic, hema ologic, ul asound, obs e-
ic-neona al, and oxic a iables.
An h opome ic a iables included in he s udy
we e p e-p egnancy weigh and heigh , BMI and sym-
physis- undal heigh (SFH). P e-p egnancy weigh
and heigh we e sel - epo ed and eco ded du ing he
ini ial p ena al examina ion a e en olmen . P e-p eg-
nancy BMI was calcula ed as weigh in kilog ams di i-
ded by he squa ed heigh in me e s (kg/m2). SFH was
measu ed in cen ime es wi h nonelas ic measu emen
ape om he uppe bo de o he symphysis pubis o
he op o he u e ine undus, o e e sed di ec ion.
Demog aphic a iables ga he ed du ing he s udy
we e ma e nal age, ma i al s a us, educa ion and oc-
cupa ion.
Haema ological a iables collec ed included hae-
mog am and se um e i in. They we e measu ed in
each imes e o p egnancy (<12, 24 and 34 weeks).
Ul asound a iables collec ed included bipa ie al
diame e (BPD), emu leng h (FL), and abdominal
ci cum e ence (AC) o he ul asound done in he hi d
imes e , be ween 33 and 35 weeks. They we e collec-
ed in o de o calcula e a s anda dized me hod7 used o
es ima e he bi h weigh a 40 week (EFW40w). We
used he equa ion de ised by Hadlock II, o ca ying
ou hei ou ine obs e ic sonog ams.
Obs e ic and neona al a iables collec ed we e pa-
i y and ges a ional age in weeks (ob ained om las
mens ual pe iod emembe ed by women). Rega ding
o he newbo n, we eco ded gende and weigh a bi h.
Toxic a iables collec ed we e p e-ges a ional o-
bacco consump ion and he numbe o ciga e es
smoked pe day in each imes e o p egnancy.
S a is ical analysis
Basic desc ip i e s a is ics a e p esen ed compa ing
p e-ges a ional BMI g oups. A e wa ds, i was ound
no mal o each o he con inuous a iables wi h he
Kolmogo o -Smi no es . The de ined le el o s a is-
ical signi icance was p<0.05.
047_8150 ¿PODEMOS MEJORAR LA PREDICCIÓN DEL PESO AL NACER? EL EFECTO DEL IMC PREGESTACIONAL NORMAL USANDO UN MODELO MULTIVARIANTE.indd 1346 18/02/15 17:18
1347
Nu Hosp. 2015;31(3):1345-1351Can we imp o e he bi h weigh
p edic ion? he e ec o no mal BMI using
a mul i a ia e model
In he bi a ia e analysis, he S uden - es was used o
compa e he means o wo quan i a i e, no malized a-
iables. Each a iable was calcula ed and compa ed be-
ween he g oup o p e-ges a ional BMI es using χ2, and
he analysis o a iance (Sche e’s hones ly signi ican
di e ences es ). In o de o es ima e he bi h weigh ,
a mul i a ia e eg ession equa ion (EFWme) using only
a iables which s a is ical signi icance, was used. Co e-
la ion be ween bo h es ima ion me hods (EFW40w and
EFWme) wi h bi h weigh we e adjus ed by ges a ional
age (38-42 weeks). Accu acy o bi h weigh es ima ion
was de e mined by calcula ing he absolu e e o o each
es ima ion me hod ([es ima ed e al weigh - ac ual bi h
weigh ] / ac ual bi h weigh ). The S uden es was used
o de e mine i his mean was signi ican ly di e en om
ze o. Di e ences be ween bo h me hods in he mean ab-
solu e e o we e assessed by he pai ed es . The mean
e o ep esen s he sum o he posi i e (o e es ima ion)
and nega i e (unde es ima ion) de ia ions om he ac-
ual bi h weigh , app oxima ing ze o in a me hod wi h
e y low o no sys ema ic e o . In o de o e alua e he
di e ence be ween EFW40w and EFWme an analysis o
he indi idual di e ences p oposed by Bland-Al man11
was used. Then bias (mean absolu e e o ) and p ecision
(SD pe cen age e o ) we e ob ained.
Da a we e analyzed using he S a is ical Package
o Social Sciences (SPSS Inc., Chicago, IL), Ve sion
15.0, and Analyse-i 3.7.
Resul s
A o al o 140 p egnan women we e app oached o
inclusion in he s udy.
A compa ison o demog aphic and clinical a iables
among he ou g oups showed signi ican di e ences
in occupa ion, ype o wo k, social s a us, and pa i y
( able I).
Table I
Demog aphic and clinical cha ac e is ics o he s udy sample by p e-p egnancy body mass index
Va iable UW (n=10) NW (n=95) OW (n=30) OB (n=5) To al (n=140) p- alue
n%n%n%n%n%
Ma e nal Age (yea s)
<25
26-29
30-34
>35
Mean (S. D.)
4 40.0 13 13.7 4 13.3 0 0.0 21 15.0 0.365*
3 30.0 28 29.5 5 16.7 1 20.0 37 26.4
3 30.0 39 41.1 16 53.3 3 60.0 61 43.6
0 0.0 15 15.8 5 16.7 1 20.0 21 15.0
29.91 (4.62)
Ma i al S a us
Ma ied
Single 7 70.0 78 82.1 25 83.3 5 100 115 82.1 0.548*
3 30.0 17 17.9 5 16.7 0 0.0 25 17.9
Educa ion
8 h g ade
High school
Uni e si y
5 50.0 33 34.7 14 46.7 3 60.0 55 39.3 0.544*
4 40.0 36 37.9 11 36.7 2 40.0 53 37.9
1 10.0 26 27.4 5 16.7 0 0.0 32 22.9
Occupa ion
Employed
Unemployed 3 30.0 74 77.9 15 50.0 2 40.0 94 67.1 0.001*
7 70.0 21 22.1 15 50.0 3 60.0 46 32.9
P e-ges a ional weigh (Kg)
Mean
(S. D.) 10 48.4
(4.49) 95 57.3
(6.60) 30 68.6
(7.41) 5 93.2
(8.40) 140 60.4
(10.7) 0.30*
Pa i y
0
³1 7 70.0 59 62.1 6 20.0 1 20.0 73 52.1 0.000**
3 30.0 36 37.9 24 80.0 4 80.0 67 47.9
Bi h weigh (g)
Mean
(S. D.) 10 3176.50
(456.63) 95 3216.37
(451.39) 30 3343.67
(507.62) 5 4153.00
(648.48) 140 3274.25
(498.95) 0.000**
S. D., s anda d de ia ion; UW, unde weigh , BMI<18.5; NW, no mal weigh , BMI: 18.5-24.9; OW, o e weigh , BMI:25-29.9; OB, obese, BMI³30.0.
* c2 es
** - es
047_8150 ¿PODEMOS MEJORAR LA PREDICCIÓN DEL PESO AL NACER? EL EFECTO DEL IMC PREGESTACIONAL NORMAL USANDO UN MODELO MULTIVARIANTE.indd 1347 18/02/15 17:18
1348 Nu Hosp. 2015;31(3):1345-1351 R Vila-Candel e al.
Mul i a ia e models o ma e nal ca ego ies UW,
OW and OB showed no s a is ically signi ican di-
e ences wi h espec o EFW40w in p edic ing bi h
weigh , and he e o e we e elimina ed.
The a iables ha showed s a is ical signi icance
wi h bi h weigh in he NW we e:
SFH 35-40 weeks (R=0.74, p<0.001), EFW40w
(R=0.63, p<0.001), GA (R=0.47, p<0.001), Fe i in
(R= -2.84, p=0.007) and numbe o ciga e es smoked
a hi d imes e (3T) (R= -2.82, p=0.006). The di e-
ence be ween he g oups (ANOVA) was no signi i-
can . Linea eg ession analysis be ween bi h weigh
and NW g oup wi h hese i e p edic o s explains i s
72% a iance. This and mul i a ia e eg ession equa-
ion a e shown in able II.
Then i was decided o s udy he di e ences o he
ac ual weigh o he newbo n be ween EFW40w and
EFWme. T- es was applied o he samples ela ed and
he di e ences we e s a is ically signi ican (p<0.001)
be ween bo h.
EFW40w, adjus ed by ges a ional age, had a co ela-
ion o 0.59 (p=0.01) a 40 weeks, and 0.69 (p=0.002)
o EFWme.
Compa ing he mean di e ences be ween
EFW40w, EFWme and bi h weigh , we obse e ha
he weigh o ul asound o e es ima es all bi h wei-
gh s. In con as , he es ima ion o he mul i a ia e
equa ion unde es ima es he bi h weigh s a weeks
38 and 41, and o e es ima es i a week 39, 40 and
42 (Figu e 1).
The di e ences in a e ages ob ained om bo h
EFW40w and EFWme wi h bi h weigh we e s a is i-
cally signi ican (p<0.001). The EFWme unde es ima-
ed bi h weigh by 0.07 g, and he EFW40w o e es-
ima ed i by 300.89 g. The e o e, p edic ion absolu e
e o was 0.53% (95% CI: -2.19-1.12) compa ed o
10.12% (95% CI: 12.81-7.43).
In o de o e alua e he p edic i e model, an obse -
a ional and e ospec i e s udy was designed. F om
he ini ial one, 138 no mal weigh p egnan women
who me c i e ia we e selec ed. Nex , di e ences be-
ween EFW40w and EFWme as well as he absolu-
e e o wi h espec o bi h weigh we e calcula ed.
To e i y he consis ency o he p edic ions we used
he Bland-Al man analysis (Figu e 2). The bi h wei-
gh alues p o ided by EFW40w a e highe han he
EFWme, wi h a di e ence o 398.6 g (95% CI: 450.5-
346.7) ( able III). The a e age e o in es ima ing he
bi h weigh wi h EFWme was 1.94% (95% CI: 0.8-
30.0) unde es ima ing he esul , whe eas he ul a-
sound e o o e es ima ed he esul 10.93% (95% CI:
-8.9-12.5).
Table II
Linea eg ession analysis showing p edic o s o bi h weigh by p e-p egnancy body mass index (NW)
Adjus ed R2 0.727
Va iables BS. E. Sig.
(Cons an ) -5667.22 850.46 0.000
SFH 35-40w 109.21 15.78 0.000
FERRITIN 3T -3.35 1.561 0.035
EFW40w 0.35 0.09 0.001
GA 13.81 3.08 0.000
Ciga e es smoked 3T -20.78 7.29 0.006
No mal weigh (NW), BMI³ 18.5 bu <24.9; SFH, symphysis- undal heigh ; EFW40w, es ima ed e al weigh a e m (40 weeks); GA, ges a ional
age. B, uns anda dised eg ession coe icien ; S.E., S anda d E o o he es ima e.
Mul i a ia e eg ession equa ion= -5667.22+.35*EFW40w+109.21*SFH+13.81*GA-3.35*Fe i in 3T 20.78*Ciga e es smoked 3T.
Table III
Analysis o bi h weigh di e ences be ween mul i a ia e model and ul asound scan EFW in p egnan women (n=138)
wi h no mal BMI
Value CI (95.0 %)**
Mean o di e ences -398.61 -450.50 -346.72
SD o di e ences 308.28
Mean -2SD* -1002.84 -1091.74 -913.94
Mean +2SD* 205.61 116.71 294.51
*SD= S anda d De ia ion
**CI (95 %)= Con idence In e al s a ed a he 95% con idence le el
047_8150 ¿PODEMOS MEJORAR LA PREDICCIÓN DEL PESO AL NACER? EL EFECTO DEL IMC PREGESTACIONAL NORMAL USANDO UN MODELO MULTIVARIANTE.indd 1348 18/02/15 17:18
1349
Nu Hosp. 2015;31(3):1345-1351Can we imp o e he bi h weigh
p edic ion? he e ec o no mal BMI using
a mul i a ia e model
Discussion
In ou s udy he e we e a numbe o a iables e-
la ed o bi h weigh in he bi a ia e analysis. Those
we e subsequen ly used o cons uc he mul i a ia e
models. E en ually we ha e shown, in he NW ca e-
go y, ha he e is a s a is ically signi ican di e en-
ce in p edic ing bi h weigh when i is compa ed o
EFW40w.
In a bi a ia e o m, SFH measu ed be ween 35-40
weeks was associa ed wi h bi h weigh o he ma e -
nal ca ego y NW. I go he highes coe icien o de e -
mina ion o all he a iables s udied, e en highe han
EFW40w a he hi d imes e . Roge s e al.12 co ela-
ed SFH wi h small- o -ges a ional-age (SGA) in an s
and 73% we e de ec ed by measu ing 3 cm o e en
below he a e age in p egnancy. In no mal-weigh mo-
he , adjus ing he SFH o ges a ional week Mele e
al.13 ob ained a no mal cu e, and an SFH below 10 h
pe cen ile was ela ed o a low bi h weigh (LBW).
In con as , Buchmann e al.14 desc ibed a SFH highe
han 40 cm as associa ed wi h an inc eased numbe o
e al mac osomia, cephalo-pel ic disp opo ion and/o
shoulde dys ocia. In ou case, undal heigh measu-
ed be ween 35 and 40 weeks, and in he p esence o
he o he a iables in he mul i a ia e model, indica es
ha bi h weigh inc eases 109.21 g o e e y cen ime-
e o u e ine heigh (95% CI: 77.6-140.6).
EFW40w was associa ed wi h bi h weigh o he
ma e nal ca ego y o NW. In p esence o he o he
a iables, he coe icien o de e mina ion was highe
han he ob ained by Ben-Ha oush5. The use o ul a-
sound as a diagnos ic me hod is well documen ed5,14-17.
Depending on he o mula used, p edic ed weigh di-
e s in i s accu acy5-16. In ou mul i a ia e model, and
in he p esence o he o he adjus ed a iables, o each
g am o a ge weigh a 40 weeks in he hi d imes-
e ul asound, bi h weigh inc eases 0.35 g (95% CI:
0.15-0.54).
Ges a ional age showed s a is ically signi ican co-
ela ion wi h bi h weigh . The a e age deli e y GA
was 278 days in p imipa ous mo he s and 279 days
o mul ipa ous mo he s. Ou mul i a ia e analysis
showed ha o e e y ex a day, he e is a e al weigh
gain o 13.81 g. This is sligh ly highe han da a ob ai-
ned by Nahum e al.16, wi h 9.66 g and 9.15 g o boys
and gi ls espec i ely, bu i is lowe han Ca alho e
al.18 wi h 28.21 g.
The smoke s’ a io be o e p egnancy was 35.0%,
and 20.7% in he las imes e , simila o o he s u-
dies e iewed19-21. Ou esul s specially indica e ha
smoking du ing he hi d imes e o p egnancy, is as-
socia ed wi h bi h weigh . I is a nega i e co ela ion
whe e inc easing numbe s o ciga e es consumed de-
c eases weigh a bi h. Consequen ly, smoking du ing
he hi d imes e seems o ha e he g ea es impac
on bi h weigh . In ac , i is known ha women who
ga e up smoking in he hi d imes e ha e babies
wi h bi h weigh s simila o hose o nonsmoke s21.
This ma ches wi h ou esul s, as smoking in he i s
wo imes e s showed no s a is ical signi icance in he
adjus ed model. The newbo n wi h low bi h weigh
becomes impo an wi h his oxic habi , and he e is a
possible ela ionship wi h he child en’s heal h de e io-
a ion because o he cy o oxic e ec 22. Pe idou e al.23
desc ibed educed newbo n weigh , by 190.8 g espec-
i ely, as compa ed o he newbo ns o non-smoking
mo he s. Ou esul s a e somewha lowe : bi h weigh
is educed abou 21 g o e e y ciga e e smoked; he
a e age numbe o ciga e es smoked pe day was 5, so
o al dec ease was abou 105 g, he same esul s ob ai-
ned by Gup a e al.24.
The amoun o e i in in he hi d imes e had an
in e se ela ionship wi h bi h weigh in No mal Wei-
gh ca ego y mo he s, so ha he less e i in, he hi-
ghe he bi h weigh . In s udies e iewed, we ound
he opposi e e ec in bo h cases: high e i in le els
we e associa ed wi h p e e m bi h, LBW and p ema-
u e up u e o memb anes25,26. O he au ho s ied o
explain a possible associa ion be ween high le els o
Fig. 1.—Di e ence bi h weigh be ween EFW40w and EFWme
By No mal-Weigh o P e-P egnancy Body Mass Index (n=95).
Fig. 2.—Di e ence bi h weigh be ween EFW40w and EFWme
By No mal-Weigh o P e-P egnancy Body Mass Index (n=138).
047_8150 ¿PODEMOS MEJORAR LA PREDICCIÓN DEL PESO AL NACER? EL EFECTO DEL IMC PREGESTACIONAL NORMAL USANDO UN MODELO MULTIVARIANTE.indd 1349 18/02/15 17:18
1350 Nu Hosp. 2015;31(3):1345-1351 R Vila-Candel e al.
e i in and e al g ow h es ic ion27, a guing ha e-
i in may be a ascula esponse o bo h in ec ious and
non-in ec ious in lamma o y diseases. Fu he s udies
a e needed o con i m his.Hämäläinen e al.28 obse -
ed ha anemia and low e i in le el du ing he i s
imes e was associa ed wi h LBW, while anemia in
he second and hi d imes e was no associa ed wi h
p e e m bi h, e al loss o isk o pe ina al complica-
ions. The e ec ound in ou s udy could be explained
as a ela ion be ween deple ion o ma e nal i on s o-
es and inc ease o i on ans e o he e us, al hough
his inc ease may be limi ed29. The deple ion o i on
in he second and hi d imes e o p egnancy in he
NW ca ego y women, physiologically declines om
he i s imes e . A he same ime, i on-ca ying ca-
paci y inc eases ( ans e in), e en when he de ici is
elimina ed by o al supplemen a ion30. In ou mul i a-
ia e model, as a nega i e ela ion, o each e i in uni
ha d opped (ng/dl), he e was a gain o 3.35 g in bi h
weigh .
The limi a ions o he mul i a ia e model (NW)
ha e o do wi h he accu acy o he ul asound and he
GA a bi h, due o he es ima ed weigh , which should
be accu a e a 40 weeks. All he newbo ns aged less
han 280 days will be o e es ima ed.
Nowadays he p edic ion o he bi h weigh h ough
ul asounds (EFW40w) has an absolu e e o ha a-
ies om 6% o 12%3-5. Accu acy can be imp o ed in
wo di e en ways: i s , by con olling he limi a ions
o he echnique and second, by adding ma e nal a-
iables om he mul i a ia e model o he ul asound
measu emen . Conside ing his pa e n, e o can be
educed up o 1.9%.
We decided o implemen wha can be conside ed
a es o p edic i e alidi y, h ough he use o a mul-
i a ia e equa ion o imp o e he es ima ion o bi h
weigh in women wi h a no mal p e-ges a ional BMI.
Then, in o de o e alua e he equa ion, he model
was used o analyze he co ela ion wi h ano he di-
e en g oup o p egnan women in a e ospec i e
s udy. In his case, 138 p egnan women, belonging o
he BMI g oup o NW, and mee ing all inclusion/ex-
clusion c i e ia o he ini ial s udy, we e selec ed. The
a e age e o wi h ou mul i a ia e model unde es i-
ma ed he bi h weigh , whe eas he ul asound a he
hi d imes e o e es ima ed he esul .
Thus, we posi i ely e alua e he mul i a ia e mo-
del ob ained, and so ha we sugges he s udy has an
impo an p ac ical applica ion. The e o e, we should
con inue by ex ending his s udy o he es o he ma-
e nal p e-ges a ional BMI g oups, which showed no
s a is ical signi icance, o de elop a new model o
each one.
Conclusion
The SFH is he a iable, which mos a ec s he p e-
dic ion o weigh a bi h.
The mul i a ia e model c ea ed imp o es he ul a-
sound measu emen by 8.99%.
The accu acy o he clinical me hod mus be de e -
mined in si ua ions which can al e he e alua ion o
weigh bi h in a ypical women, and i should be s u-
died in u u e ways o in es iga ion.
Acknowledgmen s
This s udy was suppo ed by he La Ribe a Uni e -
si y Hospi al, Spain. The au ho s wish o hank all he
s udy pa icipan s o sha ing hei da a. The au ho s
decla e no con lic o in e es .
Re e ences
1. K ame MS. De e minan s o low bi h weigh : me hodological
assessmen and me a-analysis. Bull Wo ld Heal h O gan 1987;
65 (5): 663-737.
2. de Be nabe JV, So iano T, Albaladejo R, Jua anz M, Calle
ME, Ma inez D, e al. Risk ac o s o low bi h weigh : a
e iew. Eu J Obs e Gynecol Rep od Biol 2004; 116 (1):
3-15.
3. Dudley NJ.A sys ema ic e iew o he ul asound es ima ion
o e al weigh . Ul asound Obs e Gynecol 2005; 25 (1): 80-9.
4. Oli e M, McNally G, Leade L. Accu acy o sonog aphic p e-
dic ion o bi h weigh . Aus N Z J Obs e Gynaecol 2013; 53
(6): 584-8.
5. Ben-Ha oush A, Yoge Y, Hod M, Ba J. P edic i e alue o a
single ea ly e al weigh es ima e in no mal p egnancies. Eu J
Obs e Gynecol Rep od Biol 2007; 130 (2): 187-92.
6. She man DJ, A ieli S, To bin J, Siegel G, Caspi E, Buko sky
I. A compa ison o clinical and ul asonic es ima ion o e al
weigh . Obs e Gynecol 1998; 91 (2): 212-7.
7. Chauhan SO, Hend ix NW, Magann EF, Mo ison JC, Kenney
SP, De oe LD. Limi a ions o clinical sonog aphic es ima es o
bi h weigh : Expe ience wi h 1034 pa u ien s. Obs e Gyne-
col 1998; 91 (1): 72-7.
8. Mehdizadeh A, Alaqhehbadam R, Ho san H. Compa ision o
clinical e sus ul asound es ima ion o e al weigh . Am J Pe -
ina ol 2000; 17 (5): 233-6.
9. Chien PFW, Owen P, Khan KS. Validi y o ul asound es ima-
ion o e al weigh . Obs e Gynecol 2000; 95 (6): 856-60.
10. Ba ke DJ, E iksson JG, Fo sen T, Osmond C. Fe al o igins
o adul disease: s eng h o e ec s and biological basis. In J
Epidemiol 2002; 31(6): 1235-1239.
11. Ande son NG, Jolley IJ, Wells JE. Sonog aphic es ima ion o
e al weigh : compa ison o bias, p ecision and consis ency
using 12 di e en o mulae. Ul asound Obs e Gynecol 2007;
30 (2): 173-79.
12. Roge s MS, Needham PG. E alua ion o undal heigh mea-
su emen in an ena al ca e. Aus N Z J Obs e Gynaecol 1985;
25 (2): 87-90.
13. Mele E, Pe al a P, Figue as F, Eixa ch E, Coll O, Pue o B,
e al. [U e ine heigh : no mali y cu es and diagnos ic alue
o low bi h weigh ]. P og Obs e Ginecol 2005; 48 (10):
480-6.
14. Buchmann E, Thale K. A simple clinical o mula o p edic ing
e al weigh in labou a e m, de i a ion and alida ion. S A
Med J 2009; 99 (6): 457-460.
15. Lee W, De e RL, Ebe sole JD, Huang R, Blanckae K, Rome-
o R. Bi h weigh p edic ion by h ee-dimensional ul asono-
g aphy - F ac ional limb olume. J Ul asound Med 2001; 20
(12): 1283-92.
16. Nahum GG, S anislaw H. Ul asonog aphic p edic ion o e m
bi h weigh : How accu a e is i ? Am J Obs e Gynecol 2003;
188 (2): 566-74.
047_8150 ¿PODEMOS MEJORAR LA PREDICCIÓN DEL PESO AL NACER? EL EFECTO DEL IMC PREGESTACIONAL NORMAL USANDO UN MODELO MULTIVARIANTE.indd 1350 18/02/15 17:18
1351
Nu Hosp. 2015;31(3):1345-1351Can we imp o e he bi h weigh
p edic ion? he e ec o no mal BMI using
a mul i a ia e model
17. Halaska MG, Vlk R, Feldma P, H eho cak M, K cma M, Ml-
cocho a H, e al. P edic ing e m bi h weigh using ul asound
and ma e nal cha ac e is ics. Eu J Obs e Gynecol Rep od Biol
2006; 128 (1-2): 231-5.
18. Ca alho Padilha PD, Accioly E, Chagas C, Po ela E, Da Sil a
CL, Saunde s C. Bi h weigh a ia ion acco ding o ma e nal
cha ac e is ics and ges a ional weigh gain in B azilian women.
Nu Hosp 2009; 24 (2): 207-12.
19. Pena YPD, Ma inez GR, Villag asa MPS, Pe ez VC, And es
LC, Sau as MLA, e al. Socio-cul u al, obs e ic and an h o-
pome ic cha ac e is ics o newbo n child en o mo he s who
smoke in spain. An Pedia 2012; 76 (1): 4-9.
20. Cano-Se al G, Rod íguez-Sanz M, Bo ell C, Pé ez M, Sal a-
do J. [Socioeconomic inequali ies in he p o ision and up ake
o p ena al ca e]. Gac Sani 2006; 20 (1): 25-30.
21. DiF anza JR, Aligne CA, Wei zman M. P ena al and pos na al
en i onmen al obacco smoke exposu e and child en’s heal h.
Pedia ics 2004; 113 (4): 1007-15.
22. Ba ke DJ. Fe al p og amming o co ona y hea disease.
T ends Endoc inol Me ab 2002; 13 (9): 364-8.
23. Pe idou E, Panagio opoulou K, Ka souyanni K, Spanos E,
T ichopoulos D. Tobacco smoking, p egnancy es ogens, and
bi h weigh . Epidemiology 1990; 1 (3): 247-50.
24. Gup a PC, S ee idya S. Smokeless obacco use, bi h weigh ,
and ges a ional age: popula ion based, p ospec i e coho s udy
o 1217 women in Mumbai, India. BMJ 2004; 328 (7455):
1538-40B.
25. Bha ga a M, Iye PU, Kuma R, Ramji S, Kapani V, Bha ga a
SK. Rela ionship o ma e nal se um e i in wi h e al se um
e i in, bi h-weigh and ges a ion. J T op Pedia 1991; 37
(4): 149-52.
26. Jansson L, Holmbe g L, Ekman R. Va ia ion o se um e i in
in low-bi h-weigh in an s wi h ma e nal e i in, bi h-weigh
and ges a ional-age. Ac a Haema ol 1979; 62 (5-6): 273-7.
27. Banhidy F, Acs N, Puho EH, Czeizel AE. I on de iciency ane-
mia: p egnancy ou comes wi h o wi hou i on supplemen a-
ion. Nu i ion 2011; 27 (1): 65-72.
28. Hämäläinen H, Hakka ainen K, Heinonen S. Anaemia in he
i s bu no in he second o hi d imes e is a isk ac o o
low bi h weigh . Clin Nu 2003; 22 (3): 271-5.
29. Scholl TO, Hedige ML. Anemia and i on-de iciency anemia:
compila ion o da a on p egnancy ou come. Am J Clin Nu
1994; 59 (2 Suppl): 492S-500S discussion S-1S.
30. Milman N. O al i on p ophylaxis in p egnancy: no oo li le
and no oo much! J P egnancy 2012; 2012: 514345.
047_8150 ¿PODEMOS MEJORAR LA PREDICCIÓN DEL PESO AL NACER? EL EFECTO DEL IMC PREGESTACIONAL NORMAL USANDO UN MODELO MULTIVARIANTE.indd 1351 18/02/15 17:18
